Shape Classification Method Based on the Topological Perceptual Organization Theory

1. A shape classification method implemented by a computing device, comprising the steps of: S1: extracting boundary points of shapes; S2: constructing a topological space and computing a representation of the boundary points; S3: extracting global features of the shapes from the representation of the boundary points in the topological space; S4: extracting local features of the shapes from the representation of the boundary points in a Euclidean space; S5: combining the global features and the local features through adjusting weight values of the local features according to performance of the global features; S6: classifying the shapes using a combination of the global features and the local features.

3. The method of claim 1 , wherein the global features are extracted as: h ⁡ ( k ) = ∑ i = 1 n ⁢ ∑ j = i + 1 n ⁢ θ ⁡ ( i , j ) , ⁢ if ⁢ ⁢ L ⁡ ( k ) ≤ θ ⁡ ( i , j ) < U ⁡ ( k ) θ ⁡ ( i , j ) = d * ⁡ ( i , j ) d ⁡ ( i , j ) where n is a number of boundary points of a shape, L(k) and U(k) are the lower bound and upper bound of a k th bin of a histogram.

4. The method of claim 1 , wherein the local features of a shape are extracted using a scale-invariant feature transform (SIFT) algorithm.

5. The method of claim 1 , wherein the step S5 comprises: computing a matching score of global features; and using a reciprocal of the matching score of the global features as the weight value of the local features.

7. The method of claim 1 , wherein the step S6 comprises: assembling the shapes from a same category; and separating the shapes from different categories.